Non-invasive measures of multisensory cortical feedforward and feedback influences The objective of this research is to develop and apply advanced multimodal neuroimaging methods to examine how information flows between brain areas, by using crossmodal modulation of human auditory processing as a test case. A hierarchical organization of feedforward (FF) and feedback (FB) connections among primate sensory areas has been established based on anatomical and functional connectivity patterns across cortical layers. For example, information from other sensory systems could modulate sound processing in auditory cortices (AC) through direct FF inputs, lateral inputs from other sensory cortices, and/or FB effects from higher-level polymodal areas (e.g., superior temporal sulcus STS). However, the exact way each of these mechanisms contributes to perception and cognition is an important open question. A critical barrier for resolving this question has been the lack of non-invasive techniques to make detailed inferences on FF and FB influences in cortical information processing. Such techniques are also needed to achieve better tools for the diagnosis and follow-up of disorders involving abnormal FF and FB processes, including aphasia, dyslexia, or autism. Recent studies suggest that functional FF and FB influences could be indirectly inferred from the local direction of magneto- and electroencephalography (MEG, EEG) source current estimates, as well as from frequency-band specific directed functional connectivity measures. Furthermore, recent developments in ultra-high field functional magnetic resonance imaging (fMRI) make it possible sample small voxels (< 1 mm3) at different depths of cortex, potentially enabling inferences of FF and FB type laminar activation patterns. These approaches could provide critical pieces of information regarding the hierarchical role of an area among other cortical areas, something that is not available in conventional measures of cortical activation patterns. Based on these scientific premises, our Aim 1 is to combine measures of source current direction and effective connectivity derived from MEG/EEG (Subaim 1a) with intracortical depth (or ?laminar?) analyses of 7T fMRI signals recorded simultaneously with high-density EEG data (Subaim 1b). We will compare the results with predictions based on studies of non-human primate models. Our Aim 2 is to develop novel methods for examining the neuronal mechanisms of crossmodal entrainment of AC activations in humans, including an extension of the source direction analysis to oscillatory activity. To achieve this, we will combine analyses of MEG/EEG source estimates (Subaim 2a) and analyses of simultaneously acquired laminar-resolution 7T fMRI and high-density EEG data (Subaim 2b). For both Aims, we will validate our non-invasive results by using direct brain recordings from patients with epilepsy who have intracranial electrodes implanted for medical reasons. These techniques resulting from this project will significantly augment our ability to characterize cortical processes using noninvasive neuroimaging.